# CAPE: Contrastive Action-conditioned Parallel Encoding for Embodied Planning

Source: [CAPE: Contrastive Action-conditioned Parallel Encoding for Embodied Planning](<https://www.sota2.com/research/paper/6a264ca890966ade919beafa>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-09T05:00:15.114Z

Authors: Cong Chen, Haowen Wang, Zhixiang Zhang, Pei Ren, Zhengping Che

Published: 2026-06-05T14:21:44.000Z

## Abstract

Embodied agents need to predict the future consequences of candidate actions in order to plan effectively before execution. Existing visual dynamics models learn by reconstructing future visual states or rolling out dense latent representations, which spreads learning capacity across visually salient but planning-irrelevant content rather than the action-conditioned changes that drive manipulation outcomes. We propose CAPE, a Contrastive Action-conditioned Parallel Encoding framework that learns visual dynamics by distinguishing the future outcomes induced by different action sequences. Given an initial observation and a candidate action sequence, CAPE decodes the full future latent trajectory in a single forward pass and is trained with a Goal-Convergent Contrastive Objective that aligns predictions corresponding to the same future outcome while separating those corresponding to different outcomes. On real-world DROID and zero-shot transfer to RoboCasa, CAPE substantially outperforms prior baselines on future-state retrieval, offline action matching, and closed-loop planning, while notably reducing planning-time inference cost at long prediction horizons.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Closed-loop Planning on RoboCasa Rc-R](<https://www.sota2.com/research/sota/closed-loop-planning-on-robocasa-rc-r.md>): Closed-loop Planning; RoboCasa Rc-R; Success Rate: 49.

- [Offline Action Matching on DROID](<https://www.sota2.com/research/sota/offline-action-matching-on-droid.md>): Offline Action Matching; DROID; Action Score: 62.1.

- [Future state retrieval on DROID t+1 (0.4s prediction horizon)](<https://www.sota2.com/research/sota/future-state-retrieval-on-droid-t-plus-1-0-4s-prediction-horizon.md>): Future state retrieval; DROID t+1 (0.4s prediction horizon); Hit@1: 54.11.

- [Future state retrieval on DROID t+3 (1.2s prediction horizon)](<https://www.sota2.com/research/sota/future-state-retrieval-on-droid-t-plus-3-1-2s-prediction-horizon.md>): Future state retrieval; DROID t+3 (1.2s prediction horizon); Hit@1: 48.27.

- [Future state retrieval on DROID t+5 (2.0s prediction horizon)](<https://www.sota2.com/research/sota/future-state-retrieval-on-droid-t-plus-5-2-0s-prediction-horizon.md>): Future state retrieval; DROID t+5 (2.0s prediction horizon); Hit@1: 42.97.

- [Closed-loop Planning on RoboCasa Rc-Pl](<https://www.sota2.com/research/sota/closed-loop-planning-on-robocasa-rc-pl.md>): Closed-loop Planning; RoboCasa Rc-Pl; Success Rate: 28.9.

- [MPC Planning on Computational Efficiency Evaluation](<https://www.sota2.com/research/sota/mpc-planning-on-computational-efficiency-evaluation.md>): MPC Planning; Computational Efficiency Evaluation; Total Planning Time (ms): 335.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2606.07304>)

[PDF](<https://arxiv.org/pdf/2606.07304.pdf>)
